#!/usr/bin/python 
# -*- coding: utf-8  -*-
# TODO: 
#  add per paragraph readability scores
"""
Readabilty.py

(c) Dispenser, 2007
"""

import re, wikipedia
from math import sqrt
import parser
import cgi
import cgitb; cgitb.enable()

textFiles = './text/'
ArticleWordStub = {
	'en':	"This article seems to be a stub."
}
ArticleWordLarge = {
	'en':	"This article seems to be too long and probably needs to be split."
}
ArticleLowIndex = {
	'en':	"Longer words and more compound or complex sentences may make this article more interesting, more precise and less wordy."
}
ArticleHighIndex = {
	'en':	"This article seems to have too many long words and sentences for even most university graduates to easily read and understand."
}
ArticleSizeLarge = {
	'en':	"Almost certainly should be divided up"
}
ArticleSizeBig = {
	'en':	"Probably should be divided (although the scope of a topic can sometimes justify the added reading time)"
}
Languages = {
	'':	"English",
	'simple':	"Simple English",
}
Families = {
	'wikibooks':	"Wikibooks",
	'wikipedia':	"Wikipedia",
	'wikisource':	"Wikisource",
	'wikinews':	"Wikinews",
	'wiktionary':	"Wiktionary",
	'wikiquote':	"Wikiquote",
	'wikiversity':	"Wikiversity",
}


# How AWB does it
# Remove tables and templates
# Count instances using r"[a-zA-Z]+"
flags = re.UNICODE
WordCount     = re.compile(r'[\w\-\']+', flags) #should include \-
SentenceCount = re.compile(r'\w*\w\w[ \'"\])]*[\.\!\?]+[\s\'"\]()]', flags)
def printu(ustr):
	try:
		print ustr
	except:
		print ustr.encode('utf-8')

def printOptions(dict, default):
	for key in dict:
		if key == default:
			print '<option value="%s" selected="selected">%s</option>' % (key, dict[key])
		else:
			print '<option value="%s">%s</option>' % (key, dict[key])

def printWiki(text, basehref=None, linkClass='extiw'):
	# Replace math with static image
	text = re.sub(r'(?is)<math>(.*?)</math>', r'<img class="tex" alt="Math tag" src="http://upload.wikimedia.org/math/e/e/6/ee6329486d72b1cc1b8a0e2092ddd419.png" alt="" />', text)
	text = parser.parser(text, docroot = basehref, allowHtml=True)
	# Remove Reference (no good way to parse)
	text = re.sub(r'(?is)<ref[^/>]*>.*?</ref>|<ref[^>]*/>', '<sup style="color:green;">[ref]</sup>', text)
	print text.encode('utf-8')

def wikiLink(text):
	for m in re.finditer(r'\[\[(?:([^][|]*)\||)(.*?)\]\]', text):
		link = (m.group(1) or m.group(2)).replace(' ', '_').strip('_')
		text = text.replace(m.group(), r'<a href="http://%s/wiki/%s" class="extiw">%s</a>' % (hostname, link, m.group(2)))
	return text
	
import time

def profile(text):
	print '<!-- Profiling: %s: %s -->' % (time.clock(), text)
#########################################
##
#########################################
def grade2age(grade):
	return int(grade+5.3)
def rowPrint(*cells):
	print '<tr>'
	for cell in cells:
		print('<td>%s</td>' % wikiLink(str(cell)))
	print '</tr>'
def average(list):
	if list == []:
		return 0
	else:
		return float(sum(list))/float(len(list))
class ReadingNumbers(object):
  def __init__(self, IgnoreWords = []):
#IgnoreWords	= re.split(r'\W', page.title().lower())
	IgnoreWords.append(u'wikipedia')

  def populateStats(self, text):
  	self.text      = text
  	self.markedup  = '' 
	self.words     = len(WordCount.findall(text))     or 1
	self.sentences = len(SentenceCount.findall(text)) or 1
	self.ascii     = len(re.findall(r'(?u)[ -~]', text))

	profile('senetence counting')
	for s in SentenceCount.finditer(text):
		self.markedup = self.markedup.replace(s.group(), '<span style="background:#fee;">%s</span>%s' % (s.group()[:-1],s.group()[-1]))
	
	# The actual staticstics stuff, should be branched off into an object/class
	PropNouns	= []#re.findall(r'(?<!\w.[\.\n])\s+\b((?:Mc|)[A-Z][a-z\-]{7,})\b', text)
	printu("<!-- Proper nouns:\n %r\n -->" % PropNouns)
	
	letters = 0
	unrecognized = 0
	self.HardWords = 0
	self.monoWords = 0
	syllable_list	= []
	
	self.longwords = len(re.findall(r'[^\W\d]{7,}', text))

	profile('word analysis')
	VowelsCount = re.compile(r'(?i)[aeiouy]+')
	#for word in re.finditer(r'(?u)\w+', text):
	print '<!-- vowels-less words: '
	for wordre in WordCount.finditer(text):
		s = wordre.group()
		word = wordre.group()
		if s.endswith('ses'):
			s = s[:-2]
		elif len(s) > 3 and s.endswith('s'):
			s = s[:-1]
		vowels = len(VowelsCount.findall(s))
		if s.isdigit() or s.isupper():
			syllable_list.append(len(word))
		elif vowels:
			if vowels > 1 and (s.endswith('e') or s.endswith('es') or s.endswith('ed') or s.endswith('ing') or s.endswith('ies') or s.endswith('ely')):
				vowels -= 1
			letters += len(word)	# Make sure we're adding words
			syllable_list.append(vowels)
			if word in PropNouns:
				pass
			elif vowels >= 3:
				self.HardWords += 1
				self.markedup += '<span style="background:LightPink;" title="%d">%s</span> ' % (vowels, word)
				#re.sub(r'(%s)\b([^<>])' % word, r'<u>\1</u>\2', self.markedup)
			elif vowels == 1:
				self.markedup += '<span style="background:lightgreen;" title="1">%s</span> ' % word
				self.monoWords += 1
			else:
				self.markedup += '<span style="background:lightcyan;" title="2">%s</span> ' % word
				pass
		else:
			unrecognized += 1
			printu(word)
	print '-->'
	self.letters = letters
	self.syllables = sum(syllable_list)
	self.avgSyllables = average(syllable_list)
	#print syllable_list
	print '<!--'
	print 'words: %d' % self.words
	print 'unrecognized words: %d' % unrecognized
	print 'sentences: %d' % self.sentences 
	print 'Average number of syllables per word %g' % average(syllable_list)
	print 'syllables words: %d' % len(syllable_list)
	print 'No. of syllables:%d' % sum(syllable_list)
	print '-->'

  def makeIndexes(self):
  	print '<!-- <pre>'
	# Calculate stats
# Flesch Revised formula
#	self.Flesch_Ease		= 206.835 - (1.015 * self.words/ float(self.sentences)) - (84.6 * self.avgSyllables)
	print self.words/ float(self.sentences)
	print 
	print self.avgSyllables
	print -84.6 * self.syllables
	print
	print self.words
	print '/'
	print float(self.sentences)
	print '='
	print float(self.words)/ float(self.sentences)
	print
	print 206.8
	print (-84.6 * self.syllables / float(self.words))
	print (-1.015 * self.words/ float(self.sentences))
	print '_____'
	self.Flesch_Ease		= 206.8 - (84.6 * self.syllables / float(self.words)) - (1.015 * self.words/ float(self.sentences)) 
	print self.Flesch_Ease		
	print '</pre> -->'
	self.ARI				= ( 4.71 * self.letters/ float(self.words)) + ( 0.5 * float(self.words) / float(self.sentences) ) - 21.43
	self.Flesch_Kincaid	= ( 0.39 * self.words   / float(self.sentences) ) + ( 11.8 * self.avgSyllables ) - 15.59
	self.Cloeman_Liau	= ( 5.89 * self.letters/ float(self.words) ) - ( 0.3 * float(self.sentences) / float(self.words) ) - 15.8
	self.Gunning_fog		= 0.4 * ( (float(self.words) / float(self.sentences)) + 100 * (float(self.HardWords) / float(self.words)) )
	self.SMOG			= 1.0430 * sqrt(30.0 * self.HardWords / float(self.sentences) ) + 3.1291
	self.LIX            = (self.words/float(self.sentences))+((100*self.longwords)/float(self.words))
	self.Linsear        = 0

	self.fry_x = (self.avgSyllables*100) - 108
	x = 100*float(self.sentences)/float(self.words)
	self.fry_y = -0.005*x**4 + 0.3764*x**3 - 10.611*x**2 + 136.93*x - 245.68

	self.Fry = 0


def doStats(page, raw=None):
  if not raw:
	try:
		wikiText = page.get()
		wikiText = wikiText.replace('\r\n', '\n')
	except wikipedia.NoPage:
		print '<div style="margin-left:2em;">Page is empty/non-existant</div>'
		return
	except wikipedia.IsRedirectPage:
		print '<img src="http://%s/skins-1.5/common/images/redirectltr.png" alt="#REDIRECT " /><a href="?page=%s:%s" class="redirectText">%s</a>' % (page.hostname(), page.site().language(), page._redirarg, page._redirarg)
		return
	if not wikiText.strip():
		print 'Page does not exist'
		return
	elif '{{disambig}}' in wikiText:
		printWiki(wikiText, basehref='?page=', linkClass = '')
		return
  elif raw:
  	wikiText = raw
  if True: # indent hack
  	profile('contents loaded')
	printu('<a href="http://%s/wiki/%s" class="extiw" style="font-size:200%%;">%s</a>' % (page.hostname(), page.urlname(), page.title()))
	print '<table class="wikitable" style="min-width:225px;float:left;">'
	rowPrint('<img src="" alt="" />', 
		'Paragraphs (est.)', len(re.split(r'(?um)\n\s*[\n=]', wikiText)))
	rowPrint('<img src="http://upload.wikimedia.org/wikipedia/commons/thumb/5/54/Gartoon-Galeon.png/24px-Gartoon-Galeon.png" alt="" />', 
		'Links', len(re.findall(r'\[\[[^][:]+\]\]', wikiText)))
	rowPrint('<img src="http://upload.wikimedia.org/wikipedia/commons/thumb/5/5f/Image-x-generic.svg/24px-Image-x-generic.svg.png" alt="" />',
		'Images', len(re.findall(r'\[\[ *[Ii]mage:',	wikiText)))
	rowPrint('<img src="http://upload.wikimedia.org/wikipedia/commons/thumb/c/c7/Crystal_Clear_app_virus_detected.png/24px-Crystal_Clear_app_virus_detected.png" alt="" />', 
		'Categories',	len(re.findall(r'\[\[ *[Cc]ategory:',	wikiText)))
	rowPrint('<img src="http://upload.wikimedia.org/wikipedia/commons/thumb/b/bf/Preferences-desktop-locale.svg/24px-Preferences-desktop-locale.svg.png" alt="" />', 
		'Interwiki links',	len(re.findall(r'\[\[ *[a-z\-]{2,4}:',	wikiText)))
	rowPrint('<img src="http://upload.wikimedia.org/wikipedia/commons/thumb/2/25/External.svg/16px-External.svg.png" style="padding:4px;" alt="" />', 
		'External links', len(re.findall(r'\w+://\w+\.\w+',	wikiText)))
	rowPrint('<img src="http://upload.wikimedia.org/wikipedia/commons/thumb/1/1b/X-office-document.svg/24px-X-office-document.svg.png" alt="" />', 
		'References', wikiText.lower().count('</ref>'))
	rowPrint('<img src="http://upload.wikimedia.org/wikipedia/commons/thumb/3/3b/X-office-drawing-template.svg/24px-X-office-drawing-template.svg.png" alt="" />', 
		'Templates', wikiText.lower().count('}}'))
	print '</table>'
	profile('base stats done')
	
	text = wikipedia.removeDisabledParts(wikiText)
	profile('removed disabled parts')
	# Remove links
	while re.search(r'\[\[[^][]+\]\]', text):
		text = re.sub(r'\[\[(Image|Category|[A-Za-z]{2,3}(-[A-Za-z]+)*|simple):[^][]*\]\]', '', text, re.IGNORECASE | re.DOTALL | re.UNICODE)
		text = re.sub(r'\[\[([^]|[]+\|)?([^][]+)\]\]', r'\2', text, re.IGNORECASE|re.DOTALL)
		text = re.sub(r'\[\[Image:[^][]*\]\]', '', text, re.IGNORECASE|re.DOTALL)
	profile('wikilink removal complete')

    # Remove templates	
	# Tempate striper
	stack = [text.find('{{')]
	while 10 > len(stack) > 0:
		start = stack.pop()
		if start == -1:	continue
		end   = text.find('}}', start)+2
		nextStart = text.find('{{', start+2)
		
		if end > nextStart > start:
			stack.append(start)
			stack.append(nextStart)
		elif end > start:
			text = text[:start] + text[end:]
			if stack == []:
				stack.append(text.find('{{', start))
	profile('template removal complete')

	# Remove italics and bolding
	text = re.sub(r"(?m)'''(.*?)'''", r'<b>\1</b>', text)
	text = re.sub(r"(?m)''(.*?)''",   r'<i>\1</i>', text)
	
	# Remove External links
	text = re.sub(r'\[\w+:/*[^][<>\s"]* *(.*?)\]', r'\1', text)

	# Get "readable prose"
	prose = re.compile('|'.join((
                r'\{\|.*?\|\}',
                r'<ref[^<>/]*>.*?</ref>',
                r'\{\{.*?\}\}',
                r'^=+[^\n]+=+ *$',
                r'^ *\*[^\n]*$',
                r'^ *\#[^\n]*$',
                r'^ [^\n]+$',
                r'^;+[^\n]*$',
                #r'^:+[^\n]{1,200}$', # can do full definition list
                r"^:+ *<i>\w[^\n]*$", # remove dab like lines
                r'&[^;]{2,6};')), re.IGNORECASE | re.DOTALL | re.MULTILINE | re.UNICODE).sub(r'', text)
	# Remove HTML
	text = re.sub(r'&#?\w{2,8};', r'', text)
	text = re.sub(r'</?\w+[^<>]*>', r'', text)
	prose= re.sub(r'</?\w+[^<>]*>', r'', prose)

	# Size Comparisions
	print '<table class="wikitable" style="min-width:225px;float:left;z-index:1000"><tr><td>'
	print '<div style="width:200px;"><!-- min width HACK --></div>' 
	print 'Wikitext: %#.3g KB' % (len(wikiText)/1024.)
	print '<div style="width:%dpx; height:10px; background:lightblue;"><!-- IE6 Hack --></div>' % (len(wikiText)/350)
	print 'Text: %#.3g KB' % (len(text)/1024.)
	print '<div style="width:%dpx; height:10px; background:LightCoral;"><!-- IE6 Hack --></div>' % (len(text)/350)
	print 'Readable prose: %#.3g KB' % (len(prose)/1024.)
	print '<div style="width:%dpx; height:10px; background:gold;"><!-- IE6 Hack --></div>' % (len(prose)/350)
	print '</td></tr></table>'

	textstats  = ReadingNumbers()
	textstats.populateStats(text)
	textstats.makeIndexes()
	prosestats = ReadingNumbers()
	prosestats.populateStats(prose)
	prosestats.makeIndexes()
	
	print '<table class="wikitable" style="min-width:225px;float:left;">'
	print '<tr><th></th><th>Plain text</th><th>Prose</th></tr>'
	rowPrint('Words:',              textstats.words,     prosestats.words)
	rowPrint('Sentences:',          textstats.sentences, prosestats.sentences)
	rowPrint('Letters:',            textstats.letters,   prosestats.letters)
	rowPrint('Syllables:',          textstats.syllables, prosestats.syllables)
	rowPrint('Monosyllabic words:', textstats.monoWords, prosestats.monoWords)
	rowPrint('Polysyllabic words:', textstats.HardWords, prosestats.HardWords)
	print '</table>'#'''

	print '<table class="wikitable" style="min-width:225px;float:left;">'
	print '<tr><th></th><th>Plain text</th><th>Prose</th></tr>'
	rowPrint('ASCII:','%.2f%%'%(textstats.ascii*100/float(len(textstats.text))))
	rowPrint('Characters per word:','%.2f'%(textstats.letters/float(textstats.words)), '%.2f'%(float(textstats.letters)/float(textstats.words)))
	rowPrint('Syllables per word:', '%.2f'%textstats.avgSyllables, '%.2f'%prosestats.avgSyllables)
	rowPrint('Words per sentence:', '%.2f'%(textstats.words/float(textstats.sentences)), '%.2f'%(float(prosestats.words)/float(prosestats.sentences)))
	print '</table>'#'''
	
	# Readability Indexes (print)
	print '<table class="wikitable" style="min-width:225px;float:left;">'
	print '<tr><th>Test</th><th>U.S. grade level</th><th>Reader age</th><th>Scale</th></tr>'
	rowPrint('[[Flesch Reading Ease]]',        '', '', '%.1f' % prosestats.Flesch_Ease)
	rowPrint('[[Automated Readability Index]] (ARI)','%.2f'%prosestats.ARI,            grade2age(prosestats.ARI))
	rowPrint('[[Flesch-Kincaid]]',       '%.2f'%prosestats.Flesch_Kincaid, grade2age(prosestats.Flesch_Kincaid))
	rowPrint('[[Coleman-Liau]]',         '%.2f'%prosestats.Cloeman_Liau,   grade2age(prosestats.Cloeman_Liau))
	rowPrint('[[Gunning Fog]]',          '%.2f'%prosestats.Gunning_fog,    grade2age(prosestats.Gunning_fog))
	rowPrint('[[SMOG]]',                 '%.2f'%prosestats.SMOG,           grade2age(prosestats.SMOG))
	rowPrint('[[Laesbarhedsindex]] (LIX)',     '', '', '%.2f'%prosestats.LIX)
	rowPrint('[[Linsear Write]]',              '%.2f'%prosestats.Linsear, '')
	rowPrint('[[Fry Readability Formula|Fry]]','%.2f'%prosestats.Fry, '')
	rowPrint('[[New Dale-Chall]]','', '')
	rowPrint('[[Raygor Estimate]]','', '')
	rowPrint('[[Rate Index]] (RIX)','', '')
	rowPrint('[[]]','', '')
	print '</table>'

	print '<div style="clear:both;"></div>'
	fry_left = (prosestats.fry_x*22/2+42)/3
	fry_top  = (497 - prosestats.fry_y)/3

	print '<!-- %s, %s -->' % (prosestats.fry_x, prosestats.fry_y)
	print '<table class="wikitable" style="float:left;"><tr><td>'
	print '<div style=" width:257px; height:166px; position:relative;" title="Image:Fry Graph.png">'
	print '<img src="http://upload.wikimedia.org/wikipedia/en/thumb/4/47/Fry_Graph.png/257px-Fry_Graph.png" width="257" height="166" border="0" style="position:absolute;" alt="" />'
	print '<div style="position:absolute; left:%spx; top:%spx; width:2px; height:2px;">' % (fry_left-4, fry_top-4, )
#	print '<img alt="Br.ko" src="http://upload.wikimedia.org/wikipedia/commons/thumb/2/2f/Red_pog2.svg/8px-Red_pog2.svg.png" border="0" height="8" width="8"></div>'
	print '<span style="color:red">&#x2022;</span></div>'
	print '</div>'
	print '</td></tr></table>'

	# wORD tOTALS
	print '<table class="wikitable">'
	print '<tr><td>'
	print 'words, McAlpine EFLAW miniwords, 6+ characters, 3+ sllables words, Spache unfamiliar words, DC unfamiliar words, Monosyllablic words'

	print '</td></tr>'
	print '</table>'

	profile('Suggestion')	
	# Suggestion
	if textstats.words < 300:
		suggest = ArticleWordStub
	elif textstats.words > 10000:
		suggest = ArticleWordLarge
	elif textstats.SMOG <= 8:
		suggest = ArticleLowIndex
	elif textstats.SMOG >= 16:
		suggest = ArticleHighIndex
	elif len(text) > 75e3:
		suggest = ArticleSizeLarge
	elif len(text) > 50e3:
		suggest = ArticleSizeBig
	else:
		suggest = None
	if suggest:
		print '<dl>'
		print '<dt>Notes</dt>'
		print '<dd>%s</dd>' % wikipedia.translate(page.site().language(), suggest)
		print '</dl>'

	print wikiLink('<div>The follow is what the software to recognized as "[[Wikipedia:Article length#What is and is not included as .22readable prose.22|Readable Prose]]", underlined words are polysyllabic words (syllables &#8805; 3) and highlighted words indicate punctuation.  No direct action should result from the information provided here.  Redistributed is under the terms of the [[Wikipedia:GFDL|GNU Free Documentation License]]</div>')
	
	profile('Analysis finished')

	print '<table style="clear:both; table-layout:fixed;"><tr valign="top"><td style="overflow:scroll;">'
	print '<h1>Wikitext</h1>'
	printWiki(wikiText, basehref="?page=", linkClass="")
	profile('Printed wikitext')
	print '</td><td style="overflow:scroll;">'
	print '<h1>Text</h1>'
	printWiki(text, linkClass="")
	profile('Printed text')
	print '</td><td style="width:33%;">' 
	print '<h1>Proses (Marked up)</h1>'
	printWiki(prosestats.markedup, linkClass="")
	profile('Printed prose')
	print '</td></tr></table>'
#	printu('<xmp>%s</xmp>' % text)
	
	profile('per paragraph stats')
	print """<style>
	.readgraph div{
		border:black solid 1px;
		margin:2px; 
		z-index:5;
	}
	.bar{
		background:#DAFF6B; 
		border-bottom: #aaa;
		border-right: #aaa;
		border-top: #333;
		border-left: #333;
		padding:0 4px; 
	}
	</style>"""
	## Laundry list ##
	# When conditions for readability is too short, turn the bar color red
	# Use javascript to select different test for readability
	print '<div style="margin:0.5em auto; width:40em;"><big>Per-paragraph SMOG indexes Graph</big>'
	numlist = []
	print '<div style="background:#eee; position:relative;">'
	print '<div class="readgraph" style="margin:2em; border:1px gray solid; margin:0.5em; overflow:auto; height:300px;">'
	for para in prose.split('\n\n'):
		if not para.strip(' \t\r\n:.,<>[]\'"='): continue
		pn = ReadingNumbers()
		pn.populateStats(para)
		pn.makeIndexes()
		if pn.SMOG < 4:	continue 
		color = '#DAFF6B'
		if pn.words < 40: color = "#FF6B90"
		print '<div class="bar" style="width:%sem; background:%s;" title="%s...">%.2f</div>' % (pn.SMOG, color, cgi.escape(para[:100]).encode('utf-8'), pn.SMOG)
		numlist.append(pn.SMOG)
	print '</div>'
	avg = sum(numlist)/(len(numlist) or 1)
	print '<div style="position:absolute; border:#ccc solid 1px; width:9em; height:300px; top:0; left:%sem;" title="3&sigma;, 99.7%% [&mu;-4.5, &mu;+4.5]"><br /><br/></div>' % (avg-4.5,)
	print '<div style="position:absolute; border:#888 solid 1px; width:6em; height:300px; top:0; left:%sem;" title="3&sigma;, 95.4%% [&mu;-4.5, &mu;+4.5]"><br /></div>' % (avg-3.0,)
	print '<div style="position:absolute; border:#444 solid 1px; width:3em; height:300px; top:0; left:%sem;" title="3&sigma;, 68.2%% [&mu;-1.5, &mu;+1.5]"><br/></div>' % (avg-1.5,)
	print '<div style="position:absolute; background:#DC143C; width:1px; height:300px; top:0; left:%sem;" title="Mean (&mu;) = %s"><!-- mean --></div>' % (avg, avg,)
	print '</div>'
	print """<div>
	<b>LEGEND</b> : Each bar corresponds to a paragraph.  The red line is the mean of the bars, the boxes are the population standard distribution ranges.  Thus, all the bars should lay within the boxes.  Red bars indicate that the conditions for the test have not been meet.
	</div>"""
	print '</div>'


def main():
	form = cgi.FieldStorage()
	global debug
	raw     = form.getfirst('raw', None)
	debug	= bool(form.getfirst('debug', False))
	host    = form.getfirst('hostname', 'en.wikipedia.org').split('.')
	lang    = form.getfirst('lang',     host[0])
	family  = form.getfirst('family',   host[1])
	site    = wikipedia.Site(lang, family)
	page    = wikipedia.Page(site, form.getfirst('page', ''))
	global hostname
	hostname= page.hostname()
	
	wikipedia.startContent(u'Readability for %s' % page.title(), form=False)
	print '<form action="" style="text-align:center;"><input name="page" size="40" value="%s" onchange="fixTitle(this)" />' % page.aslink()[2:-2].encode('utf-8')
#	print '<select name="lang">'
#	printOptions(Languages, page.site().language())
#	print '</select>'
#	print '<select name="family">'
#	printOptions(Families, page.site().family.name)
#	print '</select>'
	print '<input type="submit" value="Check page" /></form>'
	if page.title():
		doStats(page)
	elif raw:
		doStats(page, raw=raw.replace('_', ' '))
	else:
		wikipedia.showHelp('readability')
		
if __name__ == "__main__" and wikipedia.handleUrlAndHeader():
	try:
		main()
	finally:
		wikipedia.endContent()